The invention pertains to medical x-ray apparatus, and more particularly to an x-ray image intensifier tube of the proximity type for medical x-ray diagnostic use.
The common present day x-ray image intensifier tube is of the electrostatically focused inverter type with a 100 fold area minified output image size. This conventional inverter type x-ray image intensifier tube typically has a convexly curved, six to nine inch diameter input x-ray sensitive screen which converts the x-ray image into a light image which, in turn, is converted into electrons which are then accelerated and electrostatically focused onto an output image screen which is 100 times smaller in area than the input screen, being typically 0.6 inches to 1.0 inches in diameter. The displayed image on the output screen can be optically magnified and coupled to othe systems for radiographic or fluoroscopic purposes. Radiographic film is defined here as film which can be viewed directly without optical or electronic aids. We have found that the anatomical scale should not be minified more than 4.0 times. We found that 1.5 to 4.0 minification is acceptable. For example, for radiographic purposes, the image is optically coupled to a film camera or a photographic film. For fluoroscopic puposes, the image can be displayed either by using a system of mirrors and lenses for direct viewing or by using a closed circuit television camera and monitor for remote viewing.
The conversion efficiency of such a conventional image intensifier system is usually around 350,000 to 700,000 erg/cm.sup.2 -R or about 50,000 to 100,000 cd-sec/m.sup.2 -R, which is about 5,000 to 10,000 times the conversion efficiency of the old-time fluoroscopic screen. Part of this intensification is obtained as true electronic gain, which is about 50 to 100 times over the old-time fluoroscopic screen. Another factor of 100 gain is obtained through the 100 fold area minification of the image of the output screen.
The image quality of the conventional inverter type image intensifier tube is reasonably adequate for fluoroscopic use, but is far short of the requirement for radiographic use. The requirements for radiographic use are established by the conventional film-screen system, which demands a 20% modulation transfer function response at between 2 to 3 line pairs per millimeter.
Such conventional film-screen systems are commercially available in speeds ranging from 250 R.sup.-1 to 8000 R.sup.-1. The speed is defined as the reciprocal of the x-ray exposure in terms of roentgens, R, to the film-screen system to result in a net optical density of 1.0 on the processed film. The spatial resolving ability of the film-screen system is generally inversely proportional to the speed of the system. That is, the higher the spatial resolving ability the lower the speed of the system.
While film-screen systems have desirable system speed qualities, they have the drawback that they require taking full size photos which are difficult to store and which are becoming increasingly more expensive due to the rising cost of the silver halide x-ray film. Also, the film cannot be monitored during exposure to control the dosage or timing.
A recent article published by C. B. Johnson in the Proceedings of the Society of Photo Optical Instrumentation Engineers, Volume 35, pages 3-8 (1973), hypothetically suggests that an x-ray sensitive proximity type image intensifier may be designed with an x-ray sensitive conversion screen on one side of a glass support and a photocathode on the other side of the glass support. However, the article gives no specifics concerning the critical parameters or what might be used as the x-ray sensitive conversion screen. How this image intensifier can be designed to result in high conversion efficiency or high resolution was also not discussed.
A proximity device using a michrochannel plate (MCP) both as the primary x-ray sensitive conversion screen and as an electron multiplication device was described by S. Balter and his associates in Radiology, Volume 110, pages 673-676 (1974), and by Manley et al. in U.S. Pat. No. 3,394,261. According to an article published by J. Adams in Advances in Electronics and Electron Physics, Volume 22A (Academic Press, 1966), pages 139-153, this type of device has a very low quantum detection efficiency in the practical medical diagnostic x-ray energy range of 30-100 Kev. The device gain of the Balter article was first reported to be 20-30 cd-sec/m.sup.2 -R which is too low to be useful as a radiographic or fluoroscopic device. A higher gain device described in the same Balter article exhibited excessive noise. There is a real question whether a practical self-supporting MCP plate with uniform gain can be constructed with current technology to sizes beyond five to six inches in diameter which is not of sufficient size to produce an output useful for radiographic purposes.
Another approach involving proximity design was taken by I. C. P. Millar and his associates and their results were published in (1) IEEE Transactions on Electron Devices, Volume ED-18, pages 1101-1108 (1971), and (2) Advances in Electronics and Electron Physics, Volume 33A, pages 153-165 (1972).
Millar's approach again involves the use of a micro-channel plate (MCP). In this device, however, the MCP is used purely as an electron multiplication device and not as an x-ray conversion screen. The conversion factor for Millar's tube is reported to be around 200,000 cd-sec/m.sup.2 -R, which is above or higher than needed for fluoroscopic purposes, but is far too high for radiographic purposes. However, the output brightness of Millar's tube also exhibits strong dependence on the photocathode current density. At around a photocathode current density of 5.times.10.sup.-11 amperes/cm.sup.2 or at the equivalent x-ray input dose rate of around 0.6.times.10.sup.-3 R/sec, the output brightness of the tube starts to become sublinear in response with respect to the input x-ray dose rate. The sublinear response becomes worse at higher x-ray dose rate. This undesirable feature reduces contrast discrimination during fluoroscopy and is virtually useless for radiography. Again, it is unknown whether a large format beyond six inches in diameter, self-supporting and with uniform gain, MCP can be fabricated.
The Millar proximity type image intensifier tube has a glass envelope and an inwardly concave, titanium input window. The window is described as being 0.3 mm thick. Materials such as titanium, aluminum and beryllium cause undesirable scattering of the x-rays which reduces the image quality. Furthermore, because of the relatively high porosity and low tensile strength properties of such materials, they cannot be made as thick as desirable to maximize their x-ray transmissive properties. Still another problem with tubes constructed with such materials for the input window and glass for the tube envelope is in joining the window to the tube envelope. The materials have such dissimilar thermal expansion properties, among other differences, as to preclude their practical commercial use in a large format device.
In all such prior art x-ray image intensification devices there is the further problem of x-ray back scatter at the output display screen due to x-rays passing both out of the tube output window and coming into the tube through the output window. This can distort the displayed image and pose a danger to the user of the device.